That model is effectively what we've done with DISPATCH WG that started in RAI area over 10 years ago. For us, I think it made a big difference. I've been meaning to run stats, but we've opened and more importantly completed work in a number of WGs than we had in the past with SIPPING WG, and we progressed a number of things that had narrow interest as AD sponsored. After IETF-103 and before IETF-104, we got a new WG chartered. It does take longer to get things going in cases where we have folks new to IETF. In general, we've not had long presentations and always try to give enough time on the agenda for discussion, as that is the focus. We also have earlier deadlines for topics to allow time for discussion on the mailing list, so that the focus of discussions can be on issues and points raised on the mailing list. There's nothing to preclude any WG from not allowing agenda time for something that hasn't had any discussion on the mailing list. Or, put those last on the agenda as time allows.
Personally, I think DISPATCH has worked well because when we charter WGs, they have a very tight scope. That may not be appropriate for all areas/WGs, although, I think it's a good objective.
I also think (as usual) we're overthinking this. Chairs have a lot of leverage to optimize how they run their WGs. When I chaired the CLUE WG, we had weekly (or maybe bi-weekly) calls to focus on specific issues. We made a lot more progress in the WG with that model. We didn't do those as interims - they were just design team calls that were announced on the mailing list with minutes posted and issues tracked.
Regards,
Mary
On Wed, May 15, 2019 at 2:40 PM Michael Richardson <mcr+ietf@xxxxxxxxxxxx> wrote:
Keith Moore <moore@xxxxxxxxxxxxxxxxxxxx> wrote:
> * There should be some way of evaluating this experiment to see how well it
> works (for both WG members and others). e.g. run a survey of both WG list
> members and participants and see how well it worked for them. (Two things I
> will wonder are whether these meetings were productive for those who
> attended, and also whether the meetings were sufficiently visible and
> accessible for those who didn't attend.)
That implies we have to work out what the goals are.
Let me take a stab at this:
1) make it easier to bring new work to the IETF.
2) permit deeper and faster review of the new work, providing better
feedback.
3) reduce presentations from costly in-person meetings
It would be nice to have a before/after analysis. A key metric I'd want to
have is time between individual-00 draft and adoptedwg-00 document.
A different metric would be how many non-authors are engaged prior to
and during adoption call.
--
Michael Richardson <mcr+IETF@xxxxxxxxxxxx>, Sandelman Software Works
-= IPv6 IoT consulting =-
